• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Audio Analyzers - HP 8903B vs. QuantAsylum QA403

I’m pretty sure my thread was what w3jn was referring to. I too, was thinking about going with an all in one unit, like one of the Panasonic units, but the Quantasylum was so much cheaper and equally as impressive as measurements go. I did have a few issues very early on with the QA403, but those had everything to do with my method of measurement, which means I would’ve had the same issues had I gone with a Panasonic. Now that I’m used to using the QA403 I absolutely love it. Again, we are talking about distortion numbers that are not even audible, so really any distortion analyzer would work if you want to make sure that it’s lower than anything audible. Doing it over again I would have probably picked the Panasonic had I been able to pick it up at the same price, but they’re generally 2000 - 3000, being as that the QA403 is 1/4 to 1/5 that price I would pick the Quantasylum every day of the week. It’s a fantastic piece of gear and I use it anytime I’m about to start and anytime I finish a piece.

Dan
 
Last edited:
Indeed, your experience is the one I was thinking of, Dan. Very glad you got that working to your expectations and you're happy with it.

I also have a Rohde and Schwarz UPL, which is an all-in-one with its own screen and (DOS-based) internal computer. It has the digital audio option, including AES optical or balanced in/out. Got it cheaply as it was broken.... a shorted SMD capacitor and a fuse and it was up and working. It's much easier to run sweeps etc than on the 8903A, but I still reach for the 8903A when I'm tracking down an issue. The HP is just more intuitive to me, but then I've been using HP/Agilent equipment for most of my professional career.
 
Now somebody on eBay has HP 8903B (8 units) for $575 shipped. The item is located in Shenzhen China, no photo, but buyer gets 100% feedback rating with 1200+ sales.
 
@saabracer23 I've been reading your threads here and on QA's site. I watched the video that the Vintage Audio Review guy put out about his dummy load that he uses with the QA402. I'm trying to reconcile in my head the method he uses, versus what QA tells you to use. How is he not damaging his 402? How are you measuring safely?

Then I'm trying to figure out how the HP 8903B works with a dummy load. Add to that the comment that @ConradH made about measuring THD+N at the output of the DUT and not the dummy load (with the accompanying video from @Ray Gianelli ) and I'm now very confused.

How can one accomplish accurate measurements if you have to measure across the dummy load? How can you safely measure distortion at the DUT with a 403 or 8903B?

I'm trying to decide on an audio analyzer and I'm trying to decide between these two units. Am I understanding these things correctly? Or is the answer very obvious and I have more learning to do? Thanks all for the help!
 
These things are basically a signal generator and a "voltmeter" type device. The generator half sends the DUT (device under test) a signal. The other half, just like a voltmeter, gets attached to the output of the DUT to measure it. Like a voltmeter, it doesn't draw any significant current. It will have a range suited to the amplitude of the signal. I use a cable with a couple mini-grip clips to grab on to the wires. Google Pomona 4555-2 for an example.
 
The 8903 will display its input from the amp in volts RMS, watts, or relative dB - or calculate distortion, (S+N)/N, etc. Although the 8903's input is always high impedance and thus doesn't load down the amp, you can set the 8903 according to impedance of your dummy load so that it calculates the power correctly. Its input range is 300VRMS, which is more that 10,000 watts in an 8 ohm system.

The reason that it's preferable to attach your measurement gear directly to the output of the amp rather than at the dummy load is that there will be power losses in the wire between the amp and the dummy load. It's more than OK to measure at the dummy load if you want to account for the wire losses, but if you want to measure the power output of the amp, you measure at the amp.

Both the 403 and 8903B have differential inputs so you can isolate your measurement from chassis ground and any potential ground loops.
 
joltman said:
I'm trying to decide on an audio analyzer and I'm trying to decide between these two units. Am I understanding these things correctly? Or is the answer very obvious and I have more learning to do? Thanks all for the help!
Click to expand...
You may want to download Steven Tate's write-up on the QuantAsylum QA401:

Steven Tate's QA401 Write-up

Here's the link to the Write-up:

QA401 Write-up

Nice write-up to get someone up to speed on setting it up and using it.

Paul
 
w3jn said:
The 8903 will display its input from the amp in volts RMS, watts, or relative dB - or calculate distortion, (S+N)/N, etc. Although the 8903's input is always high impedance and thus doesn't load down the amp, you can set the 8903 according to impedance of your dummy load so that it calculates the power correctly. Its input range is 300VRMS, which is more that 10,000 watts in an 8 ohm system.
Click to expand...

The 8903B's input range being up to 300Vrms makes sense to me now. This means the dummy load is easier to make. The issue with the 8903B is it's lowest THD measurement seems to be 0.01% (per this doc). Although there are reports of getting down to 0.005%.

Paul Malloy said:
You may want to download Steven Tate's write-up on the QuantAsylum QA401:

Steven Tate's QA401 Write-up

Here's the link to the Write-up:

QA401 Write-up

Nice write-up to get someone up to speed on setting it up and using it.

Paul
Click to expand...
Paul, thanks for posting this! I hadn't seen this document before. I've read it and I'm trying to work out the best config for a dummy load for the QA403. Below is my line of thinking.

Formulas:

Vrms = 10^(dBV / 20)
Vrms = Square Root (Power Watts * System Z)
dBV = 20*Log10(Vrms/1V)
W = V^2/R
Voltage Divider: Vout = (V1*R2)/(R1+R2)

QA 403 has in input level max of 32dBV (40Vrms, 200W @ 8 Ohm). The better idea is to run it at 80% of that rating which is 25.6dBV (19.05Vrms, 45W @ 8 Ohm)

I have a 120W (into 8 ohm) power amp (Marantz 240) that I'd like to measure. So, let's do some math:

Vrms = Square Root (Power Watts * System Z)
Vrms = sqrt (120 * 8)
Vrms = 30.98

dBV = 20*Log10(30.98/1V)
dBV = 29.82

29.82 dBV is above the 80% threshold for the QA 403 (25.6dBV). So according to Steven's document, adding those two 5.6k ohm resistors in series should give us a voltage divider which removes 6dBV (half the voltage). So that would get us to ~23dBV. This method would meet my current need. What would someone do if they had a 200W amp (into 8 ohms)? Is this where the video from VAR comes in?

Using the above formula, 200W into 8 Ohms is 40Vrms which is ~32dBV. So again, too high for the 80% threshold of the QA403. If we use the same 5.6k Ohm resistors, we bring the voltage down to 20Vrms (assuming this formula works for RMS Volts). Which is slightly above the 80% limit for the QA403:

Vout = (V1*R2)/(R1+R2)
Vout = (40*5600)/(5600+5600)
Vout = 20
20Vrms = 26dBV

This is probably acceptable.

VAR is "tapping" into his resistor series to lower the dBV for his QA402 after 7 Ohms.

Vout = (40*1)/(7+1)
Vout = 5Vrms

5Vrms = 13.979dBV
32dBV - 13.9dBV = 18.1dBV (as shown in his video)

An 800W amp into 8 Ohms = 38dBV. So tapping into the -18dBV position on the dummy load would yield 20dBV, which is within the acceptable range.
The max wattage that the QA403 could handle using the VAR attenuation method would be:

25.6dBV + 18dBV = 43.6dbV
Vrms = 151.4
W (into 8 Ohms) = 2863.6

That's several high wattage resistors!

The downside of VAR's approach is that he's measuring distortion at a suboptimal location. I suppose the 8903 is doing the same thing, albeit, on it's internal attenuators. I guess there's really no way of getting around measuring high voltage power amps with "taps" (voltage divider) on the dummy load. Am I missing anything in these calcs? Is there a different way to lower the input voltage for the QA 403? It seems like the 403 is the better option for more accurate distortion measurements. I just like the thought of having a 8903B on the bench! Thanks to all who responded. Sorry for the late reply. Weekends are family time!

Jim
 
Keep in mind you're introducing a 5% error by not taking into account the 100K input impedance of the QA402, using 5.6K resistors.
 
@w3jn Thanks for bringing this up. I'm trying to figure this out. Would this be correct?

Calculate total resistance on the amp (assuming 8 Ohm load, two 5.6k resistors and 100k impedance of QA):

1/Rt = 1/R1 + 1/R2 + 1/R3

R1 = Dummy Load = 8 Ohms
R2 = 2 x 5600 Ohm resistors
R3 = 100k Impedance of QA 40x

1/Rt = 1/8 + 1/11200 + 1/100000
1/Rt = .125099
Rt = 1/.125099
Rt = 7.9936 Ohms

Thanks!
 
You are correct insofar as the impact it presents to the amp. But your voltage will be 5% low at the input of the analyzer.

Consider one of the 5.6K resistors as being in parallel with the 100K input impedance of the analyzer. The effective impedance is now (5.6K in parallel with 100K) 5.3K, and your 2:1 voltage divider is now a 2.05:1 voltage divider.
 
painttoad said:
i did some quick research. the prologix is over double the price.
Click to expand...

I use a Prologix GPIB-USB controller with my Rohde & Schwarz UPA distortion analyzer. Purchased it many years ago, and it works very well.
I checked out the "today" pricing; Holy crap, I don't think I paid even half that for mine :oops:
 
Pulled the trigger and ordered the QA403. Saelig is the way to go with getting the best price:

Saelig QA403

Free shipping, it was $615 total (tax was $25).

From Quantasylum, it would have been $660 (tax was $55). Tax was a big difference.
 
Just got notified, the QA403s are on back order. Saelig says stock will be available in a few days.
 
Last edited:
ConradH said:
FWIW, I like a traditional analyzer because it has a THD residual output that I can look at on a scope. I actually find that more useful than anything else when doing things like design or biasing an amp. Don't know if the QA403 can do that.
Click to expand...

It has a scope like amplification of the THD residual, and unlike a scope there is no noise component in the residual view. At really low THD levels the 'scope' out on analogue analyzers is mostly noise.

Also, there is a harmonics breakdown view in bargraphs, like my old Panasonic VP.

And of course full FFT up to 1M points.
 
You must log in or register to reply here.
Back
Top Bottom